Our log

H3 reimagines how we understand and organize location data through a global grid of hexagons. This article reveals five surprising insights: why hexagons make better neighbors than squares, why a flawless grid requires subtle imperfections, and how this elegant system now powers diverse real-world applications beyond ride-hailing.

This article builds on Introduction to H3 and presents basic H3 operations in Python. Although the core library is written in C, there are bindings for many other languages. H3 supports efficient spatial analysis through key functions, such as converting coordinates to indexes, finding neighboring cells, and computing distances.

In October 2025, 17 DKL team members attended PyCon España in Sevilla, blending learning, networking, and in-person connection. For a fully remote team, it was a rare opportunity to share experiences, exchange ideas, and bring the energy of the Python community directly into our workflows.

Most databases see the world as rows and columns — but reality is made of connections. Graph databases redefine how we understand data, modeling not just entities but their relationships. They turn static information into living networks, revealing context, meaning, and the foundation for intelligent, connected systems.

At Deep Kernel Labs, after years on complex projects across sectors, we learned a key lesson: uncertainty isn’t an anomaly in data; it’s inherent. Ignoring it leads to error. This article shares our practices for fostering a rigorous, critical, and adaptable data culture to support high-stakes decisions.